MARINE ECOLOGY PROGRESS SERIESMar Ecol Prog Ser

Vol. 519: 61–73, 2015doi: 10.3354/meps11101

Published January 20

INTRODUCTION

Anthropogenic activities are having a major effecton the marine ecosystem. The ongoing global releaseof greenhouse gases and subsequent warming hasforced ocean temperature to rise at a rate of ~0.4°Cover the last century (Rayner et al. 2003). The planetas a whole is currently experiencing the warmestperiod over the past 100 yr, and a 2 to 4.5°C rise in

land and ocean temperature is predicted by the endof this century (IPCC 2007). This change could haveprofound impacts on phytoplankton in terms of com-munity structure, physiology and primary production(Hughes 2000, Atkinson et al. 2003, Fu et al. 2007,Feng et al. 2009, Boyd et al. 2010). Data from the continuous plankton recorder (CPR) survey in thenortheast Atlantic have shown that rising sea surfacetemperature (SST) has increased phytoplankton

*Corresponding author: ghti@pml.ac.uk

Effect of increases in temperature and nutrients on phytoplankton community structure

and photosynthesis in the western English Channel

Yuyuan Xie1,2, Gavin H. Tilstone1,*, Claire Widdicombe1, E. Malcolm S. Woodward1, Carolyn Harris1, Morvan K. Barnes1,3

1Plymouth Marine Laboratory, Prospect Place, Plymouth PL1 3DH, UK

2Present address: College of Environment and Ecology, Xiamen University, Xiamen 361005, PR China3Present address: CNRS, Laboratoire d’Oceanographie de Villefranche, Villefranche-sur-Mer, France

ABSTRACT: Anthropogenic climate change is exerting pressures on coastal ecosystems throughincreases in temperature, precipitation and ocean acidification. Phytoplankton communitystructure and photo-physiology are therefore adapting to these conditions. Changes in phyto-plankton biomass and photosynthesis in relation to temperature and nutrient concentrations wereassessed using a 14 yr dataset from a coastal station in the western English Channel (WEC). Dino-flagellate and coccolithophorid biomass exhibited a positive correlation with temperature, reachingthe highest biomass between 15 and 17°C. Diatoms showed a negative correlation with tempera-ture, with highest biomass at 10°C. Chlorophyll a (chl a) normalised maximum light-saturated pho-tosynthetic rates (PB

m) exhibited a hyperbolic response to increasing temperature, with an initiallinear increase from 8 to 11°C and reaching a plateau from 12°C. There was, however, nosignificant positive correlation between nutrients and phytoplankton biomass or PB

m, whichreflects the lag time between nutrient input and phytoplankton growth at this coastal site. The ma-jor phytoplankton groups that occurred at this site occupied distinct thermal niches, which in turnmodified PB

m. Increasing temperature and higher water column stratification were major factors inthe initiation of dinoflagellate blooms at this site. Dinoflagellate blooms during summer also co-varied with silicate concentration and acted as a tracer of dissolved inorganic nitrogen and phos-phate from river run-off, which were subsequently reduced during these blooms. The data implythat increasing temperature and high river runoff during summer will promote dinoflagellateblooms in the WEC.

KEY WORDS: Temperature · Nutrients · Eutrophication · Phytoplankton community structure ·Photosynthesis · Western English Channel

OPENPEN ACCESSCCESS

© The authors 2015. Open Access under Creative Commons byAttribution Licence. Use, distribution and reproduction are un -restricted. Authors and original publication must be credited.

Publisher: Inter-Research · www.int-res.com

Mar Ecol Prog Ser 519: 61–73, 2015

abundance in cooler regions but decreased it inwarmer regions (Richardson & Schoeman 2004).Over the past 100 yr, the western English Channel(WEC) has experienced a sequence of warming andcooling events which have caused changes in theabundance of marine phytoplankton. A 1°C rise inSST for a decade during the 1990s caused anincrease in warm-water species in this region(Hawkins et al. 2003). In the WEC over the past 15 yrfrom 1992 to 2007, there was a decrease in diatomsand Phaeocystis sp. and an increase in dinoflagel-lates and coccolithophorids (Widdicombe et al. 2010).

Against the backdrop of global warming in Euro-pean waters, eutrophication has been a major prob-lem for the coastal zone and has resulted in large andsometimes harmful phytoplankton blooms that cancause hypoxia or toxicity (Nixon 1995, Harley et al.2006, Rabalais et al. 2009). In Europe, the North Sea,Baltic Sea and eastern English Channel are hotspotsfor eutrophication that have been linked to more fre-quent blooms of nuisance species, such as Phaeocys-tis globosa (Lancelot et al. 1987, Riegman et al. 1992,Schoemann et al. 2005). The EU Water FrameworkDirective was implemented to curb these effects,although regime shifts in the phytoplankton havestill occurred, which have been linked to the rise inSST (McQuatters-Gollop et al. 2007b). Associatedwith this, changes in the jet stream arising from smalllatitudinal temperature gradients have led to morefrequent storm events and heavy rainfall, which haveintensified over the past decade (Rahmstorf &Coumou 2011, Francis & Vavrus 2012). Recent cli-mate modelling also predicts an increase in annualrainfall coupled with the rise in temperature acrossmuch of the United Kingdom by the end of this cen-tury (Jenkins et al. 2009). Localised flooding in thecatchment around the WEC in the summer of 2007led to an increase in chlorophyll a (chl a) of 8.46 µg l−1

at in-shore stations due to high nutrient loads fromriver run-off (Rees et al. 2009).

It is often difficult to de-couple temperature andnutrient effects in marine environments, as they areintrinsically linked. In this study, we investigated theeffects of variability in temperature, nitrate, phos-phate and silicate concentrations on phytoplanktoncommunity structure and photosynthesis in the WEC.We analysed medium term trends (~14 yr) in the taxonomic biomass and photosynthetic response ofphytoplankton to variations in temperature andnutrients. The results are discussed in relation to theunderlying mechanism of seasonal succession inphytoplankton community and photosynthesis andthe implication of future environmental changes.

MATERIALS AND METHODS

Study site and sampling

Stn L4 (50° 15’ N, 4° 13’ W) is situated in the WEC,15 km southwest of Plymouth, England, with a depthof ca. 50 m (Fig. 1). Sampling at the site has beenestablished for over a century, and a long-term obser-vation on a weekly basis was initiated with profilesof temperature, salinity and fluorescence in 1988.Measurements of phytoplankton counts and chl awere added in 1992, nutrients in 2000, and primaryproduction in 2009 (Southward et al. 2004, Smyth etal. 2010). Samples were collected using a Lagrangiansampling mode whereby the ship is allowed to driftwith the sampled water body once the sampling loca-tion has been reached. The tide at the site has a max-imum range of 5.4 m and a current of 0.55 m s−1 (Pin-gree 1980). The river Tamar is the main source offreshwater flowing into the WEC, with a range of 5 to140 m3 s−1 at its mouth (Uncles & Stephens 1990).Vertical profiles of temperature, salinity and fluores-cence have been measured using SeaBird SBE19+

62

Fig. 1. Location (50° 15’ N, 4° 13’ W) of Stn L4 in the western English Channel. R: river

Xie et al.: Effect of temperature and nutrients on phytoplankton photosynthesis 63

CTD since 2002. Water samples were collected indiscrete sampling bottles or carboys and returned tothe laboratory in black-out cool-boxes within 2 h ofcollection for measurement of the parameters givenbelow. The mixed layer depth (MLD) when the den-sity difference is <0.125 kg m−4 from the surface wasused. Meteorological measurements including windspeed, wind direction and rainfall have been taken athourly intervals from the roof of Plymouth MarineLaboratory since 2003 (Smyth et al. 2010).

Nutrients, chl a and taxonomic phytoplanktonbiomass

Dissolved inorganic nutrient concentrations weredetermined by gas segmented flow colorimetricanalysis using a Bran and Luebbe AutoAnalyser(model AA3). Dissolved NO3

− + NO2− were deter-

mined by the spectrophotometric methods describedby Brewer & Riley (1965). Total ammonium wasdetermined according to Mantoura & Woodward(1983). The analysis of soluble reactive phosphorus(SRP) concentration was based on the method de -scribed by Zhang & Chi (2002). Dissolved silicate(DSi) was determined using standard colorimetricmethods (Kirkwood 1989). Seawater samples werecollected directly from the CTD rosette, stored frozenuntil analysis and then equilibrated to room temper-ature prior to analysis, following the protocols givenin the GO_SHIP manual (Hydes et al. 2010). Dis-solved inorganic nitrogen (DIN) was calculated fromthe sum of the concentration of NO3

− + NO2− and

ammonium.Fluorometric analysis of chl a was conducted using

100 ml of seawater filtered through 25 mm (nominalpore size ~0.7 µm) glass-fibre filters (GF/F) andextracted in 90% acetone overnight at 4°C. Chl aconcentration was then measured on a Turner fluo-rometer using the Welschmeyer (1994) method.

Samples for the enumeration of phytoplanktonwere fixed immediately on collection with 2%Lugol’s iodine and stored in cool, dark conditionsuntil taxonomical analysis using light microscopy following the methods given by Widdicombe et al.(2010). Cell volumes were calculated using approxi-mate geometric shapes and converted to carbon biomass using the equations of Menden-Deuer &Lessard (2000). The data are plotted as 4 phytoplank-ton groups: diatoms, coccolithophorids, dinoflagel-lates and nano-eukaryotes. The dynamics in biomassof key species within some groups were also inves -tigated.

Phytoplankton photosynthetic parameters

Phytoplankton photosynthetic parameters werecalculated from photosynthesis-irradiance (P-E) curvesmeasured using linear photosynthetrons illuminatedwith 50 W tungsten halogen lamps following themethods described by Tilstone et al. (2003). For eachdepth, 15 aliquots of 70 ml seawater within polycar-bonate bottles (Nalgene) were inoculated with 5 to10 µCi of 14C-labelled bicarbonate. Incubations weremaintained at in situ temperature for a 1.5 h period,after which the samples were filtered onto GF/Funder a vacuum pressure no greater than 27 kPa.The filters were then exposed to 37% fuming hydro -chloric acid for ~12 h and immersed in 4 ml scintilla-tion cocktail for 24 h, and beta-activity was countedon a TriCarb 2910 scintillation counter (PerkinElmer).Correction for quenching was performed using theexternal standard and the channel ratio methods.Total inorganic carbon fixation within each samplewas calculated following Tilstone et al. (2003) andnormalized to chl a, and the curves were then fittedusing the equation given by Platt et al. (1980):

PB = PBs [1 − exp(−αI/PB

s)]exp(−βI/PBs) (1)

where PBs is the potential light-saturated photosyn-

thetic rate the sample could have if there was no photoinhibition, α is the light-limited slope, β is theparameter representing the reduction by photoinhi-bition, and I is irradiance. The maximum photosyn-thetic rate (PB

m) is calculated as follows:

PBm = PB

s [α/(α + β)] [β/(α + β)]β/α (2)

Statistical analysis

Time series data from Stn L4 at 10 m between Apriland September from 2000 to 2013 were used toassess the effect of natural variability in nutrients andtemperature on phytoplankton community biomassand photosynthesis. Data from winter months werenot used since nutrients are always high and phyto-plankton biomass low during these months, which isforced by reduced PAR. Including the winter datawould therefore skew any subsequent relationshipsbetween nutrients and phytoplankton biomass.Nutrient data at 10 m was only available from 2012;prior to this, nutrients were measured at the surfaceonly. We therefore assumed that the upper mixedlayer was uniform from the surface to 10 m and usedthe surface values prior to 2012. Chl a at 10 m wasavailable from 2007 to 2013. Prior to 2007, chl a at

Mar Ecol Prog Ser 519: 61–73, 2015

10 m was calculated from surface chl a using theequation log10

Chl(10) = 0.9359 × log10Chl(0) + 0.002 (n =

215, R2 = 0.845) based on the data from successiveyears. Coincident with temperature and nutrientdata, there were n = 235 data for chl a, n = 222 forbiomass of phytoplankton groups and n = 68 for PB

m.These data are presented as box and whisker plots at1°C increments in temperature and logarithmicincrements of nutrient concentrations. The height ofthe box indicates the 25th (Q1) and 75th (Q3) per-centiles; the horizontal line and solid circle inside thebox are median and mean, respectively. The whiskerlimits are defined as Q3 + (1.5 × interquartile range,IRQ) and Q1 − 1.5 × IRQ and represent the 10th and90th percentiles. The outliers outside of the whiskersare given as open circles. Correlation between 2 vari-ables using the exact temperature and nutrient data(rather than the binned data) was determined bySpearman’s non-parametric rank correlation coeffi-cient in R software (R Development Core Team2013). To avoid increasing a Type I error through therepetition of analyses, the following procedure wasconducted: when p is the Type I error, (1 − p) repre-sents the rate of not making Type I error, the productof (1 − p) is the rate of not making Type I error for alltests. Routinely, the value with the highest Type Ierror is removed until the product of (1 − p) meet thecriterion of >0.95 (and p < 0.05), >0.99 (p < 0.01) and>0.999 (p < 0.001) successively. Thus, in the text, thecorrected p values are reported. The logit regressionwas employed to examine the possibility of dominantspecies forming blooms (with a bloom criterion of>10 mg C m−3). Temperature (T), DIN, SRP and DSiwere used as predictors, and the interaction term wasused to evaluate combined affects with bloom possi-bilities derived as follows:

F(t) = F(T,DIN,SRP,DSi) = 1/[1 − exp(−β0 − β1T − β2DIN − β3SRP − β4DSi

− β5T ×DIN − β6T ×SRP − β7T ×DSi)] (3)

F(t) is the logit regression function with t as the linearcombination of predictors. The results of F(t) rangebetween 0 and 1, representing the probability for abloom to occur. Forward selection was used based onAkaike’s information criterion (AIC) by stepwise pro-cedure in R software (R Development Core Team2013) to obtain the best predictive model. The signif-icance of each coefficient (βx) was determined, andthe significance of the whole model was determinedusing a chi-squared distribution. The possibility of abloom for the species above was predicted in eachtemperature interval by using average values ofbinned DIN, SRP and DSi as inputs.

RESULTS

Natural variability in chl a, phytoplankton biomassand PB

m in relation to temperature and nutrients in the WEC

Between April and September 2000 to 2013, tem-perature varied from 7.3 to 18.2°C with the highesttemperature in mid-August. During this period,~20% of DIN was <0.1 µmol l−1, and ~30% was>1 µmol l−1 (Fig. 2A). There was a negative inverserelationship between temperature and DIN and SRPfrom 7 to 11°C, beyond which there was no obviousrelationship, except for a slight increase in SRP from16 to 18°C (Fig. 2A,B). Spikes in DIN during July andAugust at between 14 and 18°C also occurred,

64

DIN

(µm

ol l−

1 )

02468

101214 1 4 7 24 17 21 17 30 42 51 17 4

A

SR

P (µ

mol

l−1 )

0.0

0.2

0.4

0.6

0.8

1.0

1.2 1 4 7 24 17 21 17 30 42 51 17 4B

T (°C)

DS

i (µm

ol l−

1 )

7 9 11 13 15 17

0

2

4

6

8

10

12 1 4 7 24 17 21 17 30 42 51 17 4C

●●

●

●

●●

●●

●

●●●

●

●

●● ● ●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●●

●

●

●

●● ●

●

●

●

●

●● ●

● ● ● ● ●

● ●●

● ●● ●

●●

●●

●

●●

●● ●

● ●●

●● ●

●

Fig. 2. Concentrations of (A) dissolved inorganic nitrogen(DIN), (B) soluble reactive phosphorus (SRP) and (C) dis-solved silicate (DSi) as functions of temperature (T) fromApril to September 2000–2013 at Stn L4 in the western Eng-lish Channel. For the box plots, the boundaries of the boxesrepresent the 25th and 75th percentile; the solid line withinthe box is the median; the black dot is the mean; the errorbars above and below the box indicate the 10th and 90thpercentiles, and the points beyond the error bars are the out-liers. Number of samples in discrete T ranges is given at thetop of each panel. In (A), blue crosses indicate spikes in

DIN > 1 µmol l−1 during July and August

Xie et al.: Effect of temperature and nutrients on phytoplankton photosynthesis

though these do not adversely affect the binnedmedian or mean values. For the low temperaturerange (7 to 9°C), DSi was low relative to DIN, and theaverage ratio of DSi to DIN was lower than Brzezin-ski’s ratio (molar ratio of 1). At the upper limit of thetemperature range (16 to 18°C), the mean DSi con-centration was >1.5 µmol l−1, and the average ratio ofDSi to DIN was >2; at 18°C, this reached a molar ratioof 5 (Fig. 2C). There was no significant correlationbetween chl a (p > 0.05; Fig. 3A) or total phytoplank-ton biomass and temperature (p > 0.05; Fig. 3B). Thedistribution of the binned values of both chl a andtotal phytoplankton biomass were bimodal, illustrat-ing the occurrence of 2 blooms. The first occurred inspring, between 9 and 12°C. The second was in summer at temperatures of between 14 and 17°C(Fig. 3B). Numerically, diatoms and nano-eukaryoteswere the most abundant groups and contributed, onaverage 38 and 37% of the total phytoplankton bio-mass. By comparison, dinoflagellates and coccolitho-phorids contri buted 22 and 5%, respectively. Thesegroup contributions to the total biomass changedover different temperature ranges (Fig. 4E). Dinofla-gellates, for example, formed 28% of total phyto-plankton biomass from 14 to 18°C (Fig. 4E), thoughthis occasionally rose to >50% and very occasionallyto >90%. At the same temperature range, the aver-age percentage of nano- eukaryotes and diatomsdropped to 66 from 91% at lower temperatures (Fig.4E). Over the entire temperature range for nano-eukaryotes, there was no sig nificant correlation withtemperature (p > 0.05; Fig. 4A). In contrast, therewere positive linear correlations for both coccolitho-phorids (p < 0.001; Fig. 4C) and dinoflagellates (p <

0.001; Fig. 4D), and there was a negative correlationwith diatoms (p < 0.05; Fig. 4B). The contributions of diatoms to total phytoplankton biomass were higherduring the spring bloom, but de creased with temper-ature (p < 0.01; Fig. 4E). In contrast, there was a sig-nificant positive correlation between the percentageof dinoflagellates and temperature, with the propor-tion of dinoflagellates increasing with in creasingtemperature (p < 0.001, Fig. 4E).

PBm varied from 0.2 to 7.5 mg C mg chl a−1 h−1 and

exhibited a positive linear correlation with tempera-ture from 9 to 11°C and thereafter remained rela-tively constant from 12 to 17°C with a mean value of~4 mg C mg chl a−1 h−1 (Fig. 5A). The trend in PB

m

with temperature could be explained by a hyperbolicfunction, as follows:

PBm = 3.77tanh[0.46(T − 8)] (4)

where T is temperature, and R2 = 0.112. There wereno significant correlations between chl a (p > 0.05;Fig. 3C) or total phytoplankton biomass and nutrients(p > 0.05; Fig. 3D). When there were spikes in DINfrom July to August, both chl a and total phytoplank-ton biomass varied from the lowest to the highest val-ues in the binned data (Fig. 3A,B). Dinoflagellatesexhibited a significant negative correlation with DIN(p < 0.001; Fig. 6A). Diatoms showed negative corre-lations with SRP and DSi (p < 0.01 and p < 0.001,respectively; Fig. 6B,C). Coccolithophorid biomasswas significantly positively correlated with DSi (p <0.05; Fig. 6D). Over the entire nutrient range, therewas also no significant correlation between PB

m andDIN (p > 0.05; Fig. 5B) as well as SRP (p > 0.05; datanot shown) and DSi (p > 0.05; data not shown).

65

●

●

●

●

●

●

●

●

●

Chl

a (m

g m

−3 )

1

●

4

●

7

●

24

●

17

●

21

●

17

●

30

●

42

●

51

●

17

●

4

●

A

●

●

●

●

●●

●●

0

2

4

6

8

10

1245

●

36

●

90

●

38

●

26

●

C

●

● ●●●

●●●

●

7 9 11 13 15 17

T (°C)

Tota

l phy

top

lank

ton

(mg

C m

−3 )

2

●

7

●

16

●

20

●

15

●

18

●

31

●

43

●

49

●

16

●

5

●

B●

● ●● ●

●

●

< 0.1

0.1−0.30.3−1

1−3 > 3

10

102

103

0

2

4

6

8

10

12

10

102

103

DIN (µmol l−1)

43

●

40

●

84

●

34

●

21

●

D

Fig. 3. (A) Chlorophyll a (chl a) and (B)total phytoplankton biomass as a functionof temperature (T). (C) Chl a and (D) totalphytoplankton biomass as a function ofdissolved inorganic nitrogen (DIN) fromApril to September 2000 to 2013 at Stn L4in the western English Channel. Numberfor the discrete T and DIN ranges aregiven at the top of each panel. Bluecrosses in (A) and (B) represent the bio-mass of the DIN spikes in Fig. 2A; see

Fig. 2 for box plot information

Mar Ecol Prog Ser 519: 61–73, 2015

Logit regressions of phytoplankton bloom occurrences

From 2000 to 2013 in the WEC, Phaeocystis sp. accounted for 12.6% of the nano-eukaryote biomassand tended to bloom when the temperature was<12°C (Fig. 7A, Table 1). Guinardia delicatula, Eu-campia zodiacus and Chaetoceros socialis were themost abundant diatoms and accounted for 16.3, 12.4

and 10.8% of the total diatom biomass, respectively.For these species, temperature exhibited a significantinverse relationship with biomass > 10 mg C m−3

(Table 1). G. delicatula exhibited a larger temperaturerange (10 to 16°C) during bloom conditions and ahigher frequency of bloom events (Fig. 7B). In con-trast, E. zodiacus had a narrower and lower tempera-ture range (10 to 14°C) (Fig. 7C). This was even lowerfor C. socialis (<12°C; Fig. 7D). SRP showed a consis-

66

●

●

●

●

●●

●

●

1

10

102

2

●

7

●

16

●

20

●

15

●

18

●

31

●

43

●

49

●

16

●

5

●

A

●

●

2

●

7

●

16

●

20

●

15

●

18

●

31

●

43

●

49

●

16

●

5

●

C

●

●●●

●

7 9 11 13 15 1710−2

10−1

1

10

102

103

1

10

102

10−2

10−1

10−2

10−1

1

10

102

1032

●

7

●

16

●

20

●

15

●

18

●

31

●

43

●

49

●

16

●

5

●

B

●●

●

●

●●

7 9 11 13 15 17

2

●

7

●

16

●

20

●

15

●

18

●

31

●

43

●

49

●

16

●

5

●

D

Bio

mas

s (m

g C

m−

3 )

T (°C)

T (°C)

Cha

nge

(%)

7 9 11 13 15 170

20

40

60

80

100E Dinoflagellates

CoccolithophoridsDiatomsNano-eukaryotes

Fig. 4. Relationship between carbon biomass of (A)nano-eukaryotes, (B) diatoms, (C) coccolithophoridsand (D) dinoflagellates and temperature (T) from Aprilto September, 2000–2013, at Stn L4 in the westernEnglish Channel. Number in discrete T ranges is givenat the top of each panel. Linear regression lines (solid)are given in panels: (B) log10(Biomass) = −0.048T +1.654 (R2 = 0.02, p < 0.05); (C) log10(Biomass) = 0.142T −2.468 (R2 = 0.13, p < 0.001); (D) log10(Biomass) = 0.225T− 2.711 (R2 = 0.33, p < 0.001). (E) Percentage change inphytoplankton groups to total biomass in relation to

temperature. See Fig. 2 for box plot information

9 11 13 15 170

2

4

6

8

10

PB m

(mg

C m

g−1

chl a

h−

1 ) 3 9 5 7 6 4 15 12 3A

T (°C)0.1 0.3 1 3

0

2

4

6

8

10

DIN (µmol l−1)

14 15 20 8 7B

Fig. 5. (A) Chlorophyll a normal-ized light-saturated photosyn-thetic rate (PB

m) in relation totemperature (T) and (B) in rela-tion to dissolved inorganic nitro-gen (DIN) from April to Septem-ber, 2000–2013, at Stn L4 in thewestern English Channel. Thedashed line is the hyperboliccurve fit. See Fig. 2 for box plot

information

Xie et al.: Effect of temperature and nutrients on phytoplankton photosynthesis

tent significant negative effect on possible bloom oc-currences (Table 1), indicative of the reduction in SRPduring bloom events. More than half of the dinoflagel-late biomass (59.2%) during the time series was fromKarenia mikimotoi, whilst Prorocentrum cordatum ac-counted for 7.5%. Dinoflagellate blooms tended to oc-cur when temperature was >14°C. For P. cordatum,this temperature threshold was 16°C (Fig. 7E,F), andincreasing DIN also had a positive effect (Table 1).

Emiliana huxleyi biomass rarely reached 10 mg Cm−3, and there was no single environmental variablethat was significant in the logit regression (Table 1).

67

●

●

●

●●

●

●

●

●●

●

0.1 0.3 1 3

43

●

40

●

84

●

34

●

21

●

A

DIN

●

●●

●

●

●

●

●

0.1 0.3 1 3

22

●

39

●

58

●

72

●

31

●

C

DSi

●

●

●●

●

●●

●

0.03 0.1 0.3

30

●

75

●

82

●

35

●

B

SRP0.1 0.3 1 3

10−2

10−1

1

10

102

103

10−2

10−1

1

10

102

103

10−2

10−1

1

10

102

103

10−2

10−1

1

10

102

103

22

●

39

●

58

●

72

●

31

●

D

DSi

Bio

mas

s (m

g C

m−

3 )

Fig. 6. Relationship between carbon biomass of (A) dinofla-gellates and dissolved inorganic nitrogen (DIN), (B) diatomsand soluble reactive phosphate (SRP), (C) diatoms and dis-solved silicate (DSi), (D) coccolithophorids and DSi fromApril to September, 2000–2013, at Stn L4 in the westernEnglish Channel. Number in discrete DIN ranges is given at

the top of each panel. See Fig. 2 for box plot information

0.0

0.1

0.2

0.3

0.4

0.5

● ● ● ●●

●● ●● ● ●● ●●●●●●●

● ●● ●● ●● ●● ● ●●●●

● ●● ● ●● ●● ●● ●● ● ●●● ●●●●●● ●● ●● ● ● ●● ●●● ●●●●●● ● ● ●● ● ●● ● ●● ● ●●●●●● ●● ● ● ● ● ● ●● ●● ●●● ● ●● ● ● ●● ● ●●● ●● ● ●●●●● ● ● ●● ● ●● ●●● ●● ●● ●●●● ● ● ● ● ● ●●●● ●●●● ●● ●● ●● ● ● ● ●● ●●● ● ●● ●●●●● ● ● ● ● ●● ● ● ●● ●●●●●●●●●●● ● ●● ● ● ●● ●● ●● ●●● ●●●

●

●

●

●●

●

●

●●

APhaeocystis sp.

● ● ●

●●

●

● ●● ● ●● ●●●●●● ●●

● ● ●● ●● ●● ●● ● ●●●●● ● ●

●●● ●● ●● ●● ● ●●● ●●●●●● ●● ●● ●● ● ●● ●●● ●●●

●●● ● ● ●● ● ●● ● ●●

●●●●●● ● ● ● ● ● ●

●

● ●●● ●● ● ●● ● ●●

● ●

●●

● ●● ● ●●

●●● ● ● ●● ● ●

●● ●● ●● ●●●● ● ● ● ●●●

●

●●●●

●● ●●●● ● ● ●● ●●

●●●

●●

● ● ● ● ● ●●

● ●● ●●●●●

●

● ●● ●● ●● ●●●●●

●

●

●

●

●

●

● ●●●

●

●● ●●

●●

●

●

●

●

●

●●

●

●

●

BG. delicatula

0.0

0.1

0.2

0.3

0.4

0.5

F(t)

● ● ● ●●● ● ●● ●● ● ●●●

●●●●● ●● ● ● ●● ●● ●● ●● ● ●●●●●● ● ● ● ●● ●● ●● ●● ● ●●● ●●●●●● ●● ●● ●● ● ● ●● ●●● ●●●●●● ● ● ●● ● ●● ● ●● ● ●●●●●● ●● ● ● ● ● ● ●● ●● ●●● ●● ● ●● ● ● ●● ● ●●● ●● ● ●●●●● ● ● ●● ● ●● ●●● ●● ●● ●●●● ● ● ● ● ● ●●●● ●●●● ●● ●● ●● ● ● ● ●● ●●● ● ●● ●●●●● ● ● ● ●● ●●●●●●

●●●●● ● ● ● ●● ●● ●●● ●●●

●

●

●

●

●

●

●

●

●

CE. zodiacus

● ● ● ●●● ● ●● ●● ● ●● ●●●●●● ●● ● ● ●● ●● ●● ●● ● ●●●●●● ● ● ● ●● ●● ●● ●● ● ●●● ●●●●●● ●● ●● ●● ● ● ●● ●●● ●●●●●● ● ● ●● ● ●● ● ●● ● ●●●●●● ●● ● ● ● ● ● ●● ●● ●●● ●● ● ●● ● ● ●● ● ●●● ●● ● ●●●●● ● ● ●● ● ●● ●●● ●● ●● ●●●● ● ● ● ● ● ●●●● ●●●● ●● ●● ●● ● ● ● ●● ●●● ● ●● ●●●●● ● ● ● ●● ● ● ●● ●●●●●●●●● ● ● ● ●● ●● ●● ●●● ●●●●

●

●

●

●

DC. socialis

8 10 12 16 161214 180.0

0.1

0.2

0.3

0.4

0.5

● ● ●●

●● ● ●● ●● ●●●●●●● ●● ● ● ●● ●● ● ●

● ●●●●●● ● ● ● ● ●● ●● ●●● ●●●●●● ●● ●● ●● ● ● ● ●●●●●● ● ● ●● ● ●● ●●

●●●●● ●● ● ● ● ● ● ●● ● ●●● ●● ● ●● ● ● ●● ● ●●● ●● ● ●●●●● ● ● ●● ● ●● ●●● ●● ●● ●●●● ● ● ● ● ● ●●●● ● ●● ●● ●● ● ● ● ●● ●● ● ●● ●●●●● ● ● ● ● ●●●

● ●● ●●●●●●●●●●● ● ●● ● ● ●● ●● ●●● ●●●

●

●

●

●●

●●

●●●

●●

●

●

●●

●●

●

●

●

●

E K. mikimotoi

8 10 14 18

● ● ● ●●● ● ●● ●● ● ●● ●●●●●● ●● ● ● ●● ●● ●● ●● ● ●●●●●● ● ● ● ●● ●● ●● ●● ●●

● ●●●●●● ●● ●● ●● ● ● ●● ●●●

●●●●●● ● ● ●● ● ●● ● ●● ● ●●● ●● ● ● ● ● ● ●● ●●● ●● ● ●● ● ● ●● ● ●●● ●●

●●

●●● ● ● ●● ●

●

● ●●● ●● ●●●●

● ● ● ● ● ● ●●●● ●

●

●● ●● ●● ●● ● ● ● ●● ●●● ●●● ●●●●● ● ● ● ● ●● ● ● ●● ●●●●●●●●●●● ● ●● ● ● ●● ●● ●● ●●●

●

●●

●●

●●

●

●

●●

F P. cordatum

T (°C)Fig. 7. Bloom occurrence of dominant phytoplankton speciesin relation to temperature: Phaeocystis spp., Guinardia deli-catula, Eucampia zodiacus, Chaetoceros socialus, Kareniamikimotoi, Prorocentrum cordatum. Data given in blue indi-cate bloom events (>10 mg C m−3). The logit regression lineis given in red (see ‘Methods’ for further details). F(t) is thelogit regression with t as the linear combination of predic-tors. F(t) is derived from Eq. (3) and represents the possibility

of a bloom occurring

Phaeocystis Guinardia Eucampia Chaetoceros Karenia Prorocentrum Emiliana pouchetii delicatula zodiacus socialis mikimotoi minimum huxleyi

Intercept 9.864** 4.988* 11.033** 16.358* −9.052*** −22.997*** −2.129**T −1.024*** −0.470** −0.864** −1.546** 0.463** 1.218**DIN −0.610 0.670*SRP −18.533 −27.597* −16.076* −23.282DSi −1.994T × SRP 1.083 Res. dev. 51.7 164.3 45.3 28.0 129.6 49.3 40.3df 219 218 218 219 220 219 220p <0.001 0.003 <0.001 <0.001 <0.001 <0.001 <0.001

Table 1. Predictive model of bloom (>10 mg C m−3) occurrence of dominant species in the western English Channel as a func-tion of temperature (T), dissolved inorganic nitrogen (DIN), soluble reactive phosphate (SRP) and dissolved silicate (DSi). Res.

dev.: residual deviance; × indicates the interaction term. *p < 0.05, **p < 0.01, and ***p < 0.001

Mar Ecol Prog Ser 519: 61–73, 2015

DISCUSSION

Recent interest in climate change has spawned alarge number of studies to improve our understand-ing of how rising CO2 and temperature will influencethe fate of marine ecosystems (Boyd et al. 2010 andreferences therein). There have been few studies,however, on the effect of different nutrient regimeson marine phytoplankton communities in the contextof global warming (Agawin et al. 2000, Moss et al.2003). In this study, we investigated the response ofthe phytoplankton community to ambient variationsin temperature and nutrients in the WEC. We ob -served distinct thermal niches for the major phyto-plankton groups, with dinoflagellates and coccolitho-phorids blooming at warmer temperatures (14 to18°C) and diatoms and Phaeocystis sp. occurring at alower temperature range of 7 to 12°C (Fig. 4). Thesetemperature ranges and phytoplankton communitystructure regimes are similar to those observed in theNE Atlantic ecosystem (e.g. McQuatters-Gollop et al.2007a). We found only weak and sometimes negativecorrelations between phytoplankton biomass andambient nutrient concentrations (Figs. 3 & 6).

Effect of temperature on phytoplankton composition and photosynthesis in the WEC

Dinoflagellates and coccolithophorids favoured ahigher temperature regime in the WEC. Similar

optimum temperatures (~17 to 23°C) have beenreported at the same latitude in previous work(Thomas et al. 2012). Although diatoms can inhabita large temperature range (−1.8 to 37°C) (Suzuki &Takahashi 1995, Thomas et al. 2012), they normallyinhabit cooler, well-mixed waters (Margalef 1978,Irwin et al. 2012), which was also seen for thebloom conditions of the dominant diatoms speciesin the WEC (Table 1, Fig. 7). Temperature often hasan inverse correlation with nutrients and a positivecorrelation with increasing light conditions, espe-cially during spring. The spring bloom in the WECwas also composed of nano-eukaryotes (includingPhaeocystis sp.), which usually occurred at temper-atures <14°C. This has also been observed in otherstudies (Jahnke & Baumann 1987). According tothe ‘Margalef mandala’, the stability of the watercolumn is an important factor in controlling phyto-plankton succession (Margalef 1978). Changes inMLD had no significant influence on the diatombiomass (Fig. 8A). By contrast, dinoflagellatesshowed significantly higher biomass with low MLDand stronger stratification (Fig. 8B). Stratification isoften linked with warming of the surface ocean;however, MLD can be regionally decoupled withtemperature, which can occur at Stn L4 (p > 0.05,Fig. 8C). Thus, our analysis showed that increasingdinoflagellate biomass was firstly associated with apositive increase in temperature, and secondly, thedecrease in MLD had a significant additive effect(p < 0.001).

68

● ●

●

●

●●

0 5 10 15 20 25 30

19

●

45

●

57

●

41

●

22

●

12

●

Dia

tom

s (m

g C

m−

3 ) A

●

●

●

●●

●

●

●●

● ●

●●

●

0 5 10 15 20 25 30

10−2

10−1

1

10

102

103

10−2

10−1

1

10

102

103 19

●

45

●

57

●

41

●

22

●

12

●

Din

ofla

gella

tes

(mg

C m

−3 )

B

MLD (m)

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●●●

●

● ●

●

●

●●●●

●

●

●

●

●

● ●

●

●●

●

●

●

●

●

● ●

●

●

●

● ●●

●●

●

●

●

●

●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●● ●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●

●●

●●

●

● ●

●

●

●

●

●

●

●

●

●

●

●●

●

● ●

●

●●

●●

●

●

●

●

●

●

●

●

●

●

●

●

●

● ●

●

●

●

●

●

●● ●

●

● ●●●

● ●

●●

●

●

●

●

●

●●

●

●

●

●

●

●●

●

●

●

●

●

●

●

●

●●

●

●

●●

●

● ●

●

●

●

●

● ●

●

●

●

●●

8 10 12 14 16 18

0

5

10

15

20

25

30

T (°C)

MLD

(m)

C

Fig. 8. The carbon biomass of (A) diatomsand (B) dinoflagellates in relation to mixedlayer depth (MLD); (C) MLD in relation totemperature (T). The linear regressionline (solid) is given on panel (B) aslog10(Dinoflagellates) = −0.0327MLD +0.880 (R2 = 0.07, p < 0.001). See Fig. 2 for

box plot information

Xie et al.: Effect of temperature and nutrients on phytoplankton photosynthesis

Optimal growth temperatures and activation ener-gies indicate how phytoplankton respond to thechanges in temperature (Chen et al. 2014). In theWEC, key diatom species were adapted to relativelylower temperatures (Fig. 7B,D), which resulted in theinverse relationship between diatom biomass andtemperature (Fig. 4B). By comparison, Karenia miki-motoi occurred at higher temperatures (Fig. 7E) andwas then succeeded by Prorocentrum cordatum.Such species succession may be indicative of a com-petitive advantage at increasing global tempera-tures, which can act as a key driver to the seasonalphytoplankton community succession. This has alsobeen observed in other regions such as the NorthAtlantic (McQuatters-Gollop et al. 2007a) and theScotian Shelf, where changes in phytoplankton com-munity structure closely follow temperature, whichin turn modifies PB

m (Bouman et al. 2005).Long-term observation of variability in PB

m atcoastal sites is one of the principal factors in explain-ing the seasonal variability in primary production(Gallegos 2014). PB

m is often described as a functionof temperature, typically as an Eppley curve or sev-enth-order polynomial (Eppley 1972, Cote & Platt1983, Behrenfeld & Falkowski 1997). In the WEC, wefound that a hyperbolic function between PB

m andtemperature explained a significant proportion of thevariance, with increasing PB

m at lower temperaturethen reaching a plateau at higher temperature. Bycontrast, in the Scotian Shelf, an Eppley temperaturedependency of PB

m gave a better fit (Bouman et al.2005). The differences probably arise because thelowest temperature during spring in the WEC was~7°C, and below this value, the relationship betweenPB

m and temperature may assume the more typicalEppley type curve. The intercept of 8 given in Eq. (4),therefore, arises from the fact that the temperaturedoes not go below 7°C from spring to autumn at StnL4 and is region-specific. By contrast, in the easternCanadian Arctic, where the temperature range is −2to 9°C, PB

m has been observed to follow a similar seg-mented response, with a stable upper envelope of2.0 mg C mg−1 chl a h−1 at a temperature >0°C (Li etal. 1984), suggesting different temperature depend-encies in different domains or regions. Since PB

m is arate normalized to chl a, the paralleled changes incarbon fixation and cellular content of pigmentscould result in a constant PB

m over a specific temper-ature range. Similar characteristics have beenobserved in diatoms such as Phaeodactylum tricor-nutum in culture studies (Li & Morris 1982). In theWEC, although phytoplankton species acclimate todifferent temperature regimes (Fig. 7, Table 1), the

relationship between PBm and temperature could be

modified by species succession. The linear increasein PB

m from 7 to 11°C is therefore also accompaniedby a change between diatoms and nano-eukaryotes,and the plateau corresponds to an increase in dino-flagellate and coccolithophorid biomass. A similarscenario has been observed in the neighbouring cen-tral WEC, where high PB

m was accompanied withincreasing pico-phytoplankton biomass (Napoléon etal. 2014). Although the upper limit of PB

m over thetemperature range in the plateau is quite constant,cellular chl a:carbon ratio and growth rates may in -crease with temperature when nutrients are replete(Geider 1987). From the WEC time series data, thisrelationship was not evident.

Effect of nutrients on phytoplankton compositionand photosynthesis in the WEC

Photosynthesis is regulated by nutrients, which,when replete, give rise to the spring bloom as daylength increases from winter to spring (Platt et al.1992). Nutrients determine the photosynthetic effi-ciency and growth rates of phytoplankton (Sosik &Mitchell 1994) since light-saturated rates of photo-synthesis are regulated by the synthesis of chl a,which is coupled to nitrate assimilation (Geider et al.1998). Models of photosynthetic rates have thereforebeen developed based on nutrient status of phyto-plankton under replete or limited conditions (Geideret al. 1998, Behrenfeld et al. 2002). These studiessuggest that when nutrients are depleted, photosyn-thetic rates are low, and under replete conditions,PB

m or optimum photosynthetic rates in crease. Thisshould also be coupled with temperature, which canco-vary with DIN, SRP and DSi over specific temper-ature ranges (Fig. 2). Previous work has shown, how-ever, that when nutrients are replete, further in creasesin temperature have little effect on the phytoplank-ton biomass (Moss et al. 2003). From time series datain the WEC using nutrient concentrations alone wewere unable to establish any tangible relationshipswith phytoplankton community structure. In theWEC from April to September, high DIN occursunder 2 different scenarios (Fig. 2A). The first is inspring, with high DIN resulting from previous wintermixing (Smyth et al. 2010). The second occurs insummer, when there can be occasional spikes in DINas a result of high rainfall and therefore river run-off(Rees et al. 2009). The influence of these can bedetected in the phytoplankton biomass as peaks inboth chl a and total biomass during the spring at tem-

69

Mar Ecol Prog Ser 519: 61–73, 2015

peratures of 9 to 10°C and during summer, whentemperatures are between 14 and 16°C (Fig. 3A,B).These peaks in phytoplankton biomass do not cor -relate directly with ambient nutrient concentrations(Fig. 6), as there is evidently a lag between highnutrient concentrations and phytoplankton growth.Agawin et al. (2000) and Maranón et al. (2012) havealso shown that ambient nutrient concentrations maybe uncoupled from changes in phytoplankton bio-mass. Nutrients can be high when phytoplanktonbiomass is low (e.g. during winter when nutrients arehigh, but there is light limitation). By comparison, atthe peak of a bloom when phytoplankton biomassis high, DIN, SRP or DSi may be used up rapidly andtherefore be low. Phytoplankton biomass and nutri-ents are out of phase, resulting from the lag responsebetween injection of nutrients into the euphotic zoneand phytoplankton growth. This can be further com-plicated by the fact that the nutrient supply mayoften be sporadic and not constant or continuous.From nutrient concentrations alone, it is thereforedifficult to formulate robust relationships betweenphytoplankton community structure and photo -synthesis.

The ‘irradiance-nitrate trade-off’ (Harrison et al.1990, Irwin et al. 2012) concept suggested that (1)diatoms live in high turbulence, high nutrient envi-ronments, such as cool and well-mixed waters, and(2) dinoflagellates are favoured in low turbulence,low nutrient environments. Similar to concept (1),Paerl et al. (2006) observed abrupt changes in thephytoplankton community structure during summerhurricanes, when diatoms responded to high riverrunoff and a lower residence time in the water col-umn and outcompeted dinoflagellates. In the WEC,Rees et al. (2009) also showed that during episodichigh river runoff in summer, the centric diatom

Chaetoceros debilis dominated and could outcom-pete dinoflagellates. Other studies have shown thatnano-phytoplankton often exhibit an initial responseto nutrient enhancement that can occur in upwellingsystems, which is then followed by diatoms (Tilstoneet al. 1999). In our analysis in the WEC, diatoms didnot respond to high summer temperatures evenwhen silicate was high (Figs. 2C & 9A, Table 1), butthese periods were dominated by dinoflagellates(Fig. 9B, Table 1). In contrast, laboratory studies haveshown that diatoms favour high Si:N ratios (Sommer1994). Dinoflagellate blooms in the WEC during sum-mer occurred at warmer temperatures during highstratification (Figs. 4 & 8). This in turn promoted ahigher photosynthetic response (Fig. 5A). Similarly inthe NE Atlantic, the frequency of dinoflagellateblooms has been strongly and positively correlatedwith salinity (from the Atlantic inflow), temperatureand wind speed (Edwards et al. 2006). The increasein dinoflagellate biomass at Stn L4 resulted in areduction of DIN and SRP, which resulted in anexcess of DSi (Fig. 9B). In contrast to observations inthe NE Atlantic, we found that in the WEC during thesummers of 2009 to 2012, when there was an in -crease in seasonal wind speeds, there was a paralleldecrease in the biomass of dinoflagellates (Fig. 10).This may be in part due to the fact that turbulencehas been shown to prevent the motile migration ofdinoflagellates, which reduces their ability to com-pete for resources (White 1976, Thomas & Gibson1990, Berdalet 1992). We also observed low biomassof dinoflagellates associated with deep MLDs (Fig.8B), resulting from the higher wind speeds (Fig. 10).The prevalent scenario in the English Channel is thatafter winter mixing and the onset of stratification inspring diatom biomass increases at varying times indifferent locations, depending on the strength of the

70

●

●●

●

●●●

●●

●

●

−2 −1 0 1 2

10

●

10

●

57

●

91

●

30

●

24

●

Dia

tom

s (m

g C

m−

3 )

A

●

●

●

●

●

●●

●●

●

−2 −1 0 1 2

10–2

10–1

1

10

102

103

10–2

10–1

1

10

102

103 10

●

10

●

57

●

91

●

30

●

24

●

Din

ofla

gella

tes

(mg

C m

−3 )B

DSi* (µmol l−1)

Fig. 9. Changes in carbon biomass of (A) diatoms and (B) dinoflagellates as a function of DSi*, defined as the difference between dissolved silicate (DSi) and dissolved inorganic nitrogen (DIN). See Fig. 2 for box plot information

Xie et al.: Effect of temperature and nutrients on phytoplankton photosynthesis

stratification and the attendant effects on diatoms insuspension (Ragueneau et al. 1996). Under theseconditions, diatoms (e.g. Rees et al. 2009) or nano-eukaryotes would dominate. In some coastal regions,increases in temperature coupled with an increase inprecipitation as a result of changes in the jet streamcan lead to higher river run-off and injection of nutri-ents into coastal regions, which in turn promotesunseasonal increases in the biomass of diatoms(Jenkins et al. 2009, Francis & Vavrus 2012). Fromour analysis, however, it was the dinoflagellates thatdominated under these conditions during summer(Fig. 7).

The temperature–size rule predicts that phyto-plankton will become smaller in an oligotrophic,warmer ocean (Finkel et al. 2010, Morán et al. 2010),which would be accompanied by a decrease of 2.5%cell volume per 1°C (Atkinson et al. 2003). If coastalor upwelling areas are included, the temperature-size rule can become reversed due to bottom-up con-trol (Agawin et al. 2000, Maranón et al. 2012). In theWEC, the correlation between high dinoflagellatebiomass and increasing temperature suggests areversal of the temperature–size relationship. Thistrend, however, may be forced by the ambient nutri-ent regime, as suggested in Fig. 9A. Thus, whennutrients are not limiting, this trend may modify thetemperature–size rule through inter-specific compe-tition and selective grazing pressure. Different eco-logical strategies will therefore aid specific species toout-compete others under certain environmentalconditions. In the light of this in the WEC, increasingtemperature and nutrients during localised stratifica-tion as a result of climate change and high precipita-tion will promote the growth of dinoflagellates (Figs.4D & 9B), which will in turn cause an increase in photosynthetic rates.

CONCLUSIONS

Analysis of ~14 yr of phytoplankton biomass andphotosynthetic parameters in relation to temperatureand nutrients was conducted at a coastal station inthe WEC. Dinoflagellate and coccolithophorid bio-mass exhibited a positive correlation with tempera-ture, whereas diatoms had a negative correlation.PB

m assumed a hyperbolic response to increasingtemperature, reaching a plateau at 12°C. There wasno correlation between nutrient levels and phyto-plankton biomass or photosynthetic rate. The majorphytoplankton groups occupied specific thermalniches, which in turn modified PB

m. The highest PBm

occurred at higher temperatures during stratificationwhen dinoflagellates bloomed. The analysis suggeststhat the dinoflagellate species such as Karenia miki-moito and Prorocentrum cordatum may have a selec-tive advantage under warmer climatic conditions.

Acknowledgements. We thank the crew of RV PlymouthQuest. We thank 3 anonymous reviewers for their com-ments, which significantly improved this manuscript. Y.X.was supported by the Chinese State Scholarship Fund tostudy in Plymouth Marine Laboratory as a joint PhD student(Grant No. 201206310058). G.H.T. was supported by theEuropean Union contract Information System on theEutrophication of our Coastal Seas (ISECA) (Contract no. 07-027-FR-ISECA) funded by INTERREG IVA 2 Mers SeasZeeën Cross-border Cooperation Programme 2007−2013.C.W., E.M.S.W. and C.H. were supported by the NERCNational Capability western English Channel Observatory.M.K.B. was funded by was supported by a NERC stu-dentship (NE/F012608/1).

LITERATURE CITED

Agawin NS, Duarte CM, Agusti S (2000) Nutrient and tem-perature control of the contribution of picoplankton tophytoplankton biomass and production. Limnol Oceanogr45: 591−600

Atkinson D, Ciotti BJ, Montagnes DJ (2003) Protists de -crease in size linearly with temperature: ca. 2.5% C−1.Proc R Soc B 270: 2605−2611

Behrenfeld MJ, Falkowski PG (1997) Photosynthetic ratesderived from satellite-based chlorophyll concentration.Limnol Oceanogr 42: 1−20

Behrenfeld MJ, Marañón E, Siegel DA, Hooker SB (2002)Photoacclimation and nutrient-based model of light- saturated photosynthesis for quantifying oceanic primaryproduction. Mar Ecol Prog Ser 228:103–117

Berdalet E (1992) Effects of turbulence on the marine dino-flagellate Gymnodinium nelsonii. J Phycol 28: 267−272

Bouman H, Platt T, Sathyendranath S, Stuart V (2005) De pendence of light-saturated photosynthesis on tem-perature and community structure. Deep Sea Res I 52:1284–1299

Boyd PW, Strzepek R, Fu F, Hutchins DA (2010) Environ-mental control of open-ocean phytoplankton groups: now and in the future. Limnol Oceanogr 55: 1353−1376

71

●●

0

3

6

9

12

15

�●�● �●

�●�● �●

�●

�●

�●

●

●

● 1

10

102

103

�●

�●�●

�●

�●

�●

�●

�●�●

Win

d s

pee

d (m

s−

1 )

Din

ofla

gella

tes

(mg

C m

−3 )

2004 2006 2008 2010 2012

Year

Fig. 10. Annual variability in wind speed (white bars) anddinoflagellate biomass (grey-shaded bars) from July to September, 2004–2012, at Stn L4 in the western English

Channel. See Fig. 2 for box plot information

Mar Ecol Prog Ser 519: 61–73, 2015

Brewer PG, Riley JP (1965) The automatic determination ofnitrate in sea water. Deep-Sea Res Oceanogr Abstr 12: 765−772

Chen B, Liu H, Huang B, Wang J (2014) Temperature effectson the growth rate of marine picoplankton. Mar EcolProg Ser 505: 37−47

Cote B, Platt T (1983) Day-to-day variations in the spring–summer photosynthetic parameters of coastal marinephytoplankton. Limnol Oceanogr 28: 320−344

Edwards M, Johns D, Leterme S, Svendsen E, Richardson A(2006) Regional climate change and harmful algal bloomsin the northeast Atlantic. Limnol Oceanogr 51: 820–829

Eppley RW (1972) Temperature and phytoplankton growthin the sea. Fish Bull 70: 1063−1085

Feng Y, Hare CE, Leblanc K, Rose JM and others (2009)Effects of increased pCO2 and temperature on the NorthAtlantic spring bloom. I. The phytoplankton communityand biogeochemical response. Mar Ecol Prog Ser 388: 13−25

Finkel ZV, Beardall J, Flynn KJ, Quigg A, Rees TAV, RavenJA (2010) Phytoplankton in a changing world: cell sizeand elemental stoichiometry. J Plankton Res 32: 119−137

Francis JA, Vavrus SJ (2012) Evidence linking Arctic ampli-fication to extreme weather in mid-latitudes. GeophysRes Lett 39: L06801, doi: 10.1029/2012GL051000

Fu FX, Warner ME, Zhang Y, Feng Y, Hutchins DA (2007)Effects of increased temperature and CO2 on photo -synthesis, growth, and elemental ratios in marine Syne-chococcus and Prochlorococcus (Cyanobacteria). J Phy-col 43: 485−496

Gallegos CL (2014) Long-term variations in primary produc-tion in a eutrophic sub-estuary. I. Seasonal and spatialpatterns. Mar Ecol Prog Ser 502: 53−67

Geider RJ (1987) Light and temperature dependence of thecarbon to chlorophyll a ratio in microalgae and cyano-bacteria: implications for physiology and growth ofphytoplankton. New Phytol 106: 1−34

Geider RJ, MacIntyre HL, Kana TM (1998) A dynamic re -gulatory model of phytoplanktonic acclimation to light,nutrients, and temperature. Limnol Oceanogr 43: 679–694

Harley CDG, Randall AH, Hultgren KM, Miner BG and oth-ers (2006) The impacts of climate change in coastal mar-ine systems. Ecol Lett 9: 228−241

Harrison PJ, Thomason PA, Calderwood GS (1990) Effects ofnutrient and light limitation on the biochemical composi-tion of phytoplankton. J Appl Phycol 2: 45−56

Hawkins SJ, Southward AJ, Genner MJ (2003) Detection ofenvironmental change in a marine ecosystem—evidencefrom the western English Channel. Sci Total Environ 310: 245−256

Hughes L (2000) Biological consequences of global warm-ing: is the signal already apparent? Trends Ecol Evol 15: 56−61

Hydes D, Aoyama M, Aminot A, Bakker K and others (2010)Determination of dissolved nutrients (N, P, Si) in seawa-ter with high precision and inter-comparability usinggas-segmented continuous flow analysers. In: The GO-SHIP repeat hydrography manual: a collection of expertreports and guidelines. IOCCP Rep No 14, ICPO Publ SerNo 134, v 1, 2010. UNESCO/IOC, www.go-ship.org/Hydro Man.htm

IPCC (2007) Climate Change 2007: the physical sciencebasis. Contribution of Working Group I to the FourthAssessment Report of the Intergovernmental Panel onClimate Change. Cambridge University Press, Cambridge

Irwin AJ, Nelles AM, Finkel ZV (2012) Phytoplankton nichesestimated from field data. Limnol Oceanogr 57: 787−797

Jahnke J, Baumann MEM (1987) Differentiation betweenPhaeocystis pouchetii (Har.) Lagerheim and Phaeocystisglobosa Scherffel. Hydrobiol Bull 21: 141−147

Jenkins GJ, Murphy JM, Sexton DMH, Lowe JA, Jones P,Kilsby CG (2009) UK climate projections: briefing report.Met Office Hadley Centre, Exeter

Kirkwood DS (1989) Simultaneous determination of selectednutrients in seawater. Int Coun Explor Sea (ICES) CM 29

Lancelot C, Billen G, Sournia A, Weisse T and others (1987)Phaeocystis blooms and nutrient enrichment in the conti-nental coastal zones of the North Sea. Ambio 16: 38−46

Li WKW, Morris I (1982) Temperature adaptation in Phaeo-dactylum tricornutum Bohlin: photosynthetic rate com-pensation and capacity. J Exp Mar Biol Ecol 58: 135−150

Li WKW, Smith JC, Platt T (1984) Temperature response ofphotosynthetic capacity and carboxylase activity in Arc-tic marine phytoplankton. Mar Ecol Prog Ser 17: 237−243

Mantoura RFC, Woodward EMS (1983) Optimization of theindophenol blue method for the automated determina-tion of ammonia in estuarine waters. Estuar Coast ShelfSci 17: 219−224

Maranón E, Cermeno P, Latasa M, Tadonléké RD (2012)Temperature, resources, and phytoplankton size struc-ture in the ocean. Limnol Oceanogr 57: 1266−1278

Margalef R (1978) Life-forms of phytoplankton as survivalalternatives in an unstable environment. Oceanol Acta 1: 493−509

McQuatters-Gollop A, Raitsos DE, Edwards M, Attrill MJ(2007a) Spatial patterns of diatom and dinoflagellate sea-sonal cycles in the NE Atlantic Ocean. Mar Ecol Prog Ser339: 301−306

McQuatters-Gollop A, Raitsos DE, Edwards M, Pradhan Y,Mee LD, Lavender SJ, Attrill MJ (2007b) A long-termchlorophyll dataset reveals regime shift in North Seaphytoplankton biomass unconnected to nutrient levels.Limnol Oceanogr 52: 635−648

Menden-Deuer S, Lessard EJ (2000) Carbon to volume rela-tionships for dinoflagellates, diatoms, and other protistplankton. Limnol Oceanogr 45: 569−579

Morán XAG, López-Urrutia Á, Calvo-Díaz A, Li WKW(2010) Increasing importance of small phytoplankton in awarmer ocean. Glob Change Biol 16: 1137−1144

Moss B, McKee D, Atkinson D, Collings SE and others (2003)How important is climate? Effects of warming, nutrientaddition and fish on phytoplankton in shallow lakemicrocosms. J Appl Ecol 40: 782−792

Napoléon C, Fiant L, Raimbault V, Riou P, Claquin P (2014)Dynamics of phytoplankton diversity structure and pri-mary productivity in the English Channel. Mar Ecol ProgSer 505: 49−64

Nixon SW (1995) Coastal marine eutrophication: a defini-tion, social causes, and future concerns. Ophelia 41: 199−219

Paerl HW, Valdes LM, Peierls BL, Adolf JE, Harding LWJ(2006) Anthropogenic and climatic influences on theeutrophication of large estuarine ecosystems. LimnolOceanogr 51: 448−462

Pingree R (1980) Physical oceanography of the Celtic Seaand English Channel. In: Banner FT (ed) The north-westEuropean shelf seas: the sea bed and the sea in motion.II. Physical and chemical oceanography and physicalresources. Elsevier, Amsterdam p 415−465

Platt T, Gallegos C, Harrison W (1980) Photoinhibition of

72

Xie et al.: Effect of temperature and nutrients on phytoplankton photosynthesis

photosynthesis in natural assemblages of marine phyto-plankton. J Mar Res 38: 687−701

Platt T, Sathyendranath S, Ulloa O, Harrison WG, HoepffnerN, Goes J (1992) Nutrient control of phytoplankton pho-tosynthesis in the Western North Atlantic. Nature 356:229–231

R Development Core Team (2013) R: a language and en -vironment for statistical computing. R Foundation forStatistical Computing, Vienna, available at www.r-project. org

Rabalais NN, Turner RE, Díaz RJ, Justi D (2009) Globalchange and eutrophication of coastal waters. ICES J MarSci 66: 1528−1537

Ragueneau O, Queguiner B, Treguer P (1996) Contrast inbiological responses to tidally-induced vertical mixingfor two macrotidal ecosystems of Western Europe. EstuarCoast Shelf Sci 42: 645−665

Rahmstorf S, Coumou D (2011) Increase of extreme eventsin a warming world. Proc Natl Acad Sci USA 108: 17905−17909

Rayner NA, Parker DE, Horton EB, Folland CK and others(2003) Global analyses of sea surface temperature, seaice, and night marine air temperature since the late nine-teenth century. J Geophys Res D 108: 4407, doi: 10/1029/2002JD002670

Rees AP, Hope SB, Widdicombe CE, Dixon JL, WoodwardEMS, Fitzsimons MF (2009) Alkaline phosphatase activ-ity in the western English Channel: elevations inducedby high summertime rainfall. Estuar Coast Shelf Sci 81: 569−574

Richardson AJ, Schoeman DS (2004) Climate impact onplankton ecosystems in the Northeast Atlantic. Science305: 1609−1612

Riegman R, Noordeloos AAM, Cadée GC (1992) Phaeocystisblooms and eutrophication of the continental coastalzones of the North Sea. Mar Biol 112: 479−484

Schoemann V, Becquevort S, Stefels J, Rousseau V, LancelotC (2005) Phaeocystis blooms in the global ocean andtheir controlling mechanisms: a review. J Sea Res 53: 43−66

Smyth TJ, Fishwick JR, Lisa AM, Cummings DG and others(2010) A broad spatio-temporal view of the western Eng-lish Channel observatory. J Plankton Res 32: 585−601

Sommer U (1994) Are marine diatoms favoured by high Si: Nratios? Mar Ecol Prog Ser 115: 309−315

Sosik HM, Mitchell BG (1994) Effects of temperature ongrowth, light absorption, and quantum yield in Duna -liella tertiolecta (Chlorophyceae). J Phycol 30: 833–840

Southward AJ, Langmead O, Hardman-Mountford NJ,Aiken J and others (2004) Long-term oceanographic andecological research in the western English Channel. AdvMar Biol 47: 1−105

Suzuki Y, Takahashi M (1995) Growth responses of severaldiatom species isolated from various environments totemperature. J Phycol 31: 880−888

Thomas WH, Gibson CH (1990) Quantified small-scale tur-bulence inhibits a red tide dinoflagellate, Gonyaulaxpolyedra Stein. Deep-Sea Res A 37: 1583−1593

Thomas MK, Kremer CT, Klausmeier CA, Litchman E (2012)A global pattern of thermal adaptation in marine phyto-plankton. Science 338: 1085−1088

Tilstone GH, Figueiras F, Fermín E, Arbones B (1999) Signif-icance of nanophytoplankton photosynthesis and pri-mary production in a coastal upwelling system (Ría deVigo, NW Spain). Mar Ecol Prog Ser 183: 13−27

Tilstone GH, Figueiras F, Lorenzo LM, Arbones B (2003)Phytoplankton composition, photosynthesis and primaryproduction during different hydrographic conditions atthe Northwest Iberian upwelling system. Mar Ecol ProgSer 252: 89−104

Uncles R, Stephens J (1990) The structure of vertical currentprofiles in a macrotidal, partly-mixed estuary. Estuaries13: 349−361

Welschmeyer NA (1994) Fluorometric analysis of chloro-phyll a in the presence of chlorophyll b and pheopig-ments. Limnol Oceanogr 39: 1985−1992

White AW (1976) Growth inhibition caused by turbulence inthe toxic marine dinoflagellate Gonyaulax excavata.J Fish Res Board Can 33: 2598−2602

Widdicombe CE, Eloire D, Harbour D, Harris RP, SomerfieldPJ (2010) Long-term phytoplankton community dynam-ics in the western English Channel. J Plankton Res 32: 643−655

Zhang JZ, Chi J (2002) Automated analysis of nanomolarconcentrations of phosphate in natural waters with liquidwaveguide. Environ Sci Technol 36: 1048−1053

73

Editorial responsibility: Graham Savidge,Portaferry, UK

Submitted: February 7, 2014; Accepted: October 25, 2014Proofs received from author(s): December 20, 2014